Cellular and Molecular Neurophysiology

Cellular and Molecular Neurophysiology

Fourth edition

Constance Hammond

Table of Contents

Cover

Title page

Copyright

Foreword

Acknowledgments

I: Neurons: excitable and secretory cells that establish synapses

Chapter 1: Neurons

Abstract

1.1. Neurons have a cell body from which emerge two types of processes: the dendrites and the axon

1.2. Neurons are highly polarized cells with a differential distribution of organelles and proteins

1.3. Axonal transport allows bidirectional communication between the cell body and the axon terminals

1.4. Neurons connected by synapses form networks or circuits

1.5. Summary: the neuron is an excitable and secretory cell presenting an extreme functional regionalization

Chapter 2: Neuron–glial cell cooperation

Abstract

2.1. Astrocytes form a vast cellular network or syncytium between neurons, blood vessels and the surface of the brain

2.2. Oligodendrocytes form the myelin sheaths of axons in the central nervous system and allow the clustering of Na+ channels at nodes of Ranvier

2.3. Microglial cells are the macrophages of the central nervous system

2.4. Schwann cells are the glial cells of the peripheral nervous system; they form the myelin sheath of axons or encapsulate neurons

Chapter 3: Ionic gradients, membrane potential and ionic currents

Abstract

3.1. There is an unequal distribution of ions across THE neuronal plasma membrane. The notion of concentration gradient

3.2. There is a difference of potential between the two faces of the membrane, called membrane potential (Vm)

3.3. Concentration gradients and membrane potential determine the direction of the passive movements of ions through ionic channels: the electrochemical gradient

3.4. The passive diffusion of ions through an open channel creates a current

3.5. A particular membrane potential, the resting membrane potential Vrest

3.6. A simple equivalent electrical circuit for the membrane at rest

3.7. How to experimentally change Vrest

3.8. Summary

Appendix 3.1. The active transport of ions by pumps and transporters maintain the unequal distribution of ions

Appendix 3.2. The passive diffusion of ions through an open channel

Appendix 3.3. The Nernst equation

Chapter 4: The voltage-gated channels of Na+ action potentials

Abstract

4.1. Properties of action potentials

4.2. The depolarization phase of Na+-dependent action potentials results from the transient entry of Na+ ions through voltage-gated Na+ channels

4.3. The repolarization phase of the sodium-dependent action potential results from Na+ channel inactivation and partly from K+ channel activation

4.4. Sodium-dependent action potentials are initiated at the axon initial segment in response to a membrane depolarization and then actively propagate along the axon

Chapter 5: The voltage-gated channels of Ca²+ action potentials: Generalization

Abstract

5.1. Properties of Ca²+-dependent action potentials

5.2. The transient entry of Ca²+ ions through voltage-gated Ca²+ channels is responsible for the depolarizing phase or the plateau phase of Ca²+-dependent action potentials

5.3. The repolarization phase of Ca²+-dependent action potentials results from the activation of K+ currents IK and IKCa

5.4. Calcium-dependent action potentials are initiated in axon terminals and in dendrites

5.5. A note on voltage-gated channels and action potentials

Chapter 6: The chemical synapses

Abstract

6.1. The synaptic complex’s three components: presynaptic element, synaptic cleft and postsynaptic element

6.2. The interneuronal synapses

6.3. The neuromuscular junction is the group of synaptic contacts between the terminal arborization of a motor axon and a striated muscle fiber

6.4. The synapse between the vegetative postganglionic neuron and the smooth muscle cell

6.5. Example of a neuroglandular synapse

6.6. Summary

Chapter 7: Neurotransmitter release

Abstract

7.1. Observations and questions

7.2. Presynaptic processes I: from presynaptic spike to [Ca²+]i increase

7.3. Presynaptic processes II: from [Ca²+] increase to synaptic vesicle fusion

7.4. Processes in the synaptic cleft: from transmitter release in the cleft to transmitter clearance from the cleft

7.5. Summary (Figures 7.16 and  7.17)

II: Ionotropic and metabotropic receptors in synaptic transmission

Chapter 8: The ionotropic nicotinic acetylcholine receptors

Abstract

8.1. Observations

8.2. The torpedo or muscle nicotinic receptor of acetylcholine is a heterologous pentamer α2βγδ

8.3. Binding of two acetylcholine molecules favors conformational change of the protein towards the open state of the cationic channel

8.4. The nicotinic receptor desensitizes

8.5. nAChR-mediated synaptic transmission at the neuromuscular junction

8.6. Nicotinic transmission pharmacology

8.7. Summary

Chapter 9: The ionotropic GABAA receptor

Abstract

9.1. Observations and questions

9.2. GABAA receptors are hetero-oligomeric proteins with a structural heterogeneity

9.3. Binding of two GABA molecules leads to a conformational change of the GABAA receptor into an open state; the GABAA receptor desensitizes

9.4. Pharmacology of the GABAA receptor

9.5. GABAA-mediated synaptic transmission

9.6. Summary

Chapter 10: The ionotropic glutamate receptors

Abstract

10.1. There are three different types of ionotropic glutamate receptors. They have a common structure and all participate in fast glutamatergic synaptic transmission

10.2. AMPA receptors are an ensemble of cationic receptor-channels with different permeabilities to Ca²+ ions

10.3. Kainate receptors are an ensemble of cationic receptor channels with different permeabilities to Ca²+ ions

10.4. NMDA receptors are cationic-receptor-channels highly permeable to Ca²+ ions; they are blocked by Mg²+ ions at voltages close to the resting potential, which confers strong voltage dependence

10.5. Synaptic responses to glutamate are mediated by NMDA and non-NMDA receptors

10.6. Summary

Chapter 11: The metabotropic GABAB receptors

Abstract

11.1. GABAB receptors were originally discovered because of their insensitivity to bicuculline and their sensitivity to baclofen

11.2. Structure of the GABAB receptor

11.3. Summary

11.4. GABAB receptors are G-protein coupled to a variety of different effector mechanisms

11.5. Summary

11.6. The functional role of GABAB receptors in synaptic activity

11.7. Summary

Chapter 12: The metabotropic glutamate receptors

Abstract

12.1. The identification of the eight metabotropic glutamate receptor subtypes

12.2. How do metabotropic glutamate receptors carry out their function? Structure–function studies of metabotropic glutamate receptors

12.3. How to identify selective compounds acting at the metabotropic glutamate receptor – towards the development of new therapeutic drugs

12.4. What biochemical means do metabotropic glutamate receptors utilize to elicit physiological changes in the nervous system? Signal transduction studies of metabotropic glutamate receptors

12.5. How is the activity of metabotropic glutamate receptors modulated? Studies of mGluR desensitization

12.6. Metabotropic glutamate receptors modulate neuronal excitability

12.7. Metabotropic glutamate receptors mediate and modulate synaptic transmission

12.8. Pre- and postsynaptic functional assembly of metabotropic glutamate receptors

12.9. Physiological roles of metabotropic glutamate receptor – a study of knockout models

12.10. Summary

III: Somato-dendritic processing and plasticity of postsynaptic potentials

Chapter 13: Somato-dendritic processing of postsynaptic potentials I: Passive properties of dendrites

Abstract

13.1. Propagation of excitatory and inhibitory postsynaptic potentials through the dendritic arborization

13.2. Summation of excitatory and inhibitory postsynaptic potentials

13.3. Summary

Chapter 14: Subliminal voltage-gated currents of the somato-dendritic membrane

Abstract

14.1. Observations and questions

14.2. The subliminal voltage-gated currents that depolarize the membrane

14.3. The subliminal voltage-gated currents that hyperpolarize the membrane

14.4. Conclusions

Chapter 15: Somato-dendritic processing of postsynaptic potentials II. Role of sub-threshold depolarizing voltage-gated currents

Abstract

15.1. Persistent Na+ channels are present in the axo-somatic region of neocortical neurons; INaP boosts EPSPs

15.2. T-type Ca²+ channels are present in the dendrites of cortical neurons; ICaT boosts EPSPs

15.3. The hyperpolarization-activated cationic current Ih is present in dendrites of hippocampal pyramidal neurons; Ih attenuates EPSPs

15.4. A-type K+ channels are present in the dendrites of hippocampal neurons; IA attenuates EPSPs

15.5. Functional consequences

15.6. Conclusions

Chapter 16: Somato-dendritic processing of postsynaptic potentials III. Role of high-voltage-activated depolarizing currents

Abstract

16.1. High-voltage-activated Na+ and/or Ca²+ channels are present in the dendritic membrane of some CNS neurons, but are they distributed with comparable densities in soma and dendrites?

16.2. High-voltage-activated Ca²+ channels are present in the dendritic membrane of some CNS neurons, but are they distributed with comparable densities in soma and dendrites?

16.3. Functional consequences

16.4. Conclusions

Chapter 17: Firing patterns of neurons

Abstract

17.1. Medium spiny projection neurons of the striatum

17.2. Inferior olivary cells

17.3. Purkinje cells are pacemaker neurons that respond by a complex spike followed by a period of silence

17.4. Thalamic and subthalamic neurons

Chapter 18: Synaptic plasticity

Abstract

18.1. Short-term potentiation (STP) of a cholinergic synaptic response as an example of short-term plasticity: the cholinergic response of muscle cells to motoneuron stimulation

18.2. Long-term potentiation (LTP) of a glutamatergic synaptic response: example of the glutamatergic synaptic response of pyramidal neurons of the CA1 region of the hippocampus to Schaffer collateral activation

18.3. The long-term depression (LTD) of a glutamatergic response: example of the response of Purkinje cells of the cerebellum to parallel fiber stimulation

18.4. Long-term synaptic modification induced by relative timing between pre- and postsynaptic activity: spike-timing-dependent plasticity (STDP)

18.5. The homeostatic plasticity of a glutamatergic response: example of the synaptic scaling at neocortical gluTaMatergic synapses

IV: The hippocampal network

Chapter 19: The adult hippocampal network

Abstract

19.1. Observations and questions

19.2. The hippocampal circuitry

19.3. Activation of interneurons evokes inhibitory GABAergic responses in postsynaptic pyramidal cells

19.4. Activation of principal cells evokes excitatory glutamatergic responses in postsynaptic interneurons and other principal cells (synchronization in CA3)

19.5. Oscillations in the hippocampal network: example of sharp waves (SPW)

19.6. Summary

Chapter 20: Maturation of the hippocampal network

Abstract

20.1. GABAergic neurons and GABAergic synapses develop prior to glutamatergic ones

20.2. GABAA receptor-mediated responses differ in developing and mature brains

20.3. Maturation of coherent networks activities

20.4. Conclusions

Contributors

Index

Copyright

Academic Press is an imprint of Elsevier

32 Jamestown Road, London NW1 7BY, UK

525 B Street, Suite 1800, San Diego, CA 92101-4495, USA

225 Wyman Street, Waltham, MA 02451, USA

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK

Copyright © 2015, 2008, 2001, 1996 Elsevier Ltd. All rights reserved

Cover figure: Ghost neuron.

Newborn rat hippocampal pyramidal neuron in a dissociated culture. Neurons were transfected with the green fluorescent protein (GFP) attached to a protein of the cell’s cytoskeleton (MAP2β) located to the cell body and dendritic arbor. Artificially colored here to evoke a ghostly ambience. Supplied by Damien Guimond, Institut de Neurobiologie de la Méditerranée, INSERM UMR 901, Marseille, France.

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology Rights

Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: permissions@elsevier.com. Alternatively, visit the Science and Technology Books website at www.elsevierdirect.com/rights for further information

Notice

No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

ISBN: 978-0-12-397032-9

For information on all Academic Press publications visit our website at http://store.elsevier.com/

Typeset by Thomson Digital

Printed and bound in United States of America

Foreword

This excellent and highly acclaimed textbook by Hammond and co-authors, now in its fourth edition, remains faithful to its central philosophy that teaching science cannot simply rely on the presentation of facts, but must also include the intellectual journey that gives birth to key discoveries and results in the solution of long-standing puzzles. Science is in constant motion, and neuroscience, in particular, cannot be understood without appreciating how ground-breaking experiments were designed to test novel hypotheses using a combination of intuition, knowledge and experience. Readers will greatly appreciate this book for several reasons. Chief among them is that the text introduces them to the core scientific process, beyond simply presenting solutions to problems. A unique aspect of this textbook is that most figures are reproduced from the original papers that first demonstrated a given finding. This is an excellent didactic approach, because these original figures not only convey information concerning particular experimental arrangements and results, but they also introduce students to the history of neuroscience first hand. At the bottom of each figure legend, for example, readers will find the full reference including the authors, title, and journal for the paper that reported the particular discovery. There is no better way to teach students integrity and self-confidence than to introduce them to the original papers.

Because the current edition contains many updated chapters and appendices, students will learn about experiments performed by pioneering authors whom they can actually meet in person at scientific conferences. This is where the true power of the book lies, as this approach enables students to familiarize themselves with the ever-changing, flesh-and-blood frontlines of cutting-edge research, while they are still in the process of studying the key concepts of molecular and cellular neuroscience. The topics range from elementary properties of excitable cells to detailed discussions of ion channels, receptors, and synaptic transmission, all the way to dendritic integration and various forms of neuronal plasticity. This book is a concise, yet in-depth, highly informative text that will continue to inspire present and future practitioners of neuroscience.

Ivan Soltesz, PhD

University of California, Irvine

Acknowledgments

Constance Hammond would like to thank her colleagues who took part of their time to give their expertise: Agnès Baude for chapters 1, 6, and 19 and Romain Nardou for chapters 14 and 19.

And the following individuals who have contributed to previous editions of the book or who helped in carefully reading the proofs:

Andrea Nistri, International School for Advanced Studies (SISSA), Trieste, Italy and Aron Gutman (deceased) formerly of Kaunas Medical Academy, Kaunas, Lithuania (previous Chapter 4, now part of Chapter 3); Gautam Bhave and Robert Gereau, Baylor College of Medicine, Houston, TX, USA (Chapter 12); Yusuf Tan, Bogazici University, Istanbul, Turkey (Appendix 5.1); Charles Bourque; Diana Ferrari (Chapter 7); and Geneviève Chazal (Chapters 19 and 20).

I

Neurons: excitable and secretory cells that establish synapses

Chapter 1: Neurons

Chapter 2: Neuron–glial cell cooperation

Chapter 3: Ionic gradients, membrane potential and ionic currents

Chapter 4: The voltage-gated channels of Na+ action potentials

Chapter 5: The voltage-gated channels of Ca2+ action potentials: Generalization

Chapter 6: The chemical synapses

Chapter 7: Neurotransmitter release

Chapter 1

Neurons

Constance Hammond

Abstract

Neurons are independent cells making specific contacts called synapses, with hundreds or thousands of other neurons sometimes greatly distant from their cell bodies. The neurons connected together form circuits, and so the nervous system is composed of neuronal networks which transmit and process information. Neurons are excitable cells. Depending on the information they receive, neurons generate electrical signals and propagate them along their processes. This capacity is due to the presence of particular proteins in their plasma membrane which allow the selective passage of ions: the ion channels. Neurons are also secretory cells. Their secretory product is called a neurotransmitter. The release of a neurotransmitter occurs only in restricted regions, the synapses. The neurotransmitter is released in the extracellular space. The synaptic secretion is highly focalized and directed specifically on cell regions to which the neuron is connected. The synaptic secretion is then different (with only a few exceptions) from other secretory cells, such as from hormonal and exocrine cells which respectively release their secretory products into the general circulation (endocrine secretion) or the external environment (exocrine secretion).

Keywords

axon

axon terminal

axonal transport

dendrite, dendritic spine

dendritic transport

dynein

excitable cell

Golgi type I neuron

Golgi type II neuron

kinesin

network

secretory cell

synapse

Outline

1.1 Neurons have a cell body from which emerge two types of processes: the dendrites and the axon 4

1.2 Neurons are highly polarized cells with a differential distribution of organelles and proteins 6

1.3 Axonal transport allows bidirectional communication between the cell body and the axon terminals 11

1.4 Neurons connected by synapses form networks or circuits 18

1.5 Summary: the neuron is an excitable and secretory cell presenting an extreme functional regionalization 22

Further reading 23

By using the silver impregnation method developed by Golgi (1873), Ramon y Cajal studied neurons, and their connections, in the nervous system of numerous species. Based on his own work (1888) and that of others (e.g. Forel, His, Kölliker and Lenhossék), he proposed the concept that neurons are isolated units connected to each other by contacts formed by their processes: ‘The terminal arborizations of neurons are free and are not joined to other terminal arborizations. They make contacts with the cell bodies and protoplasmic processes of other cellular elements.’

As proposed by Cajal, neurons are independent cells making specific contacts called synapses, with hundreds or thousands of other neurons sometimes greatly distant from their cell bodies. The neurons connected together form circuits, and so the nervous system is composed of neuronal networks which transmit and process information. In the nervous system, there is another class of cells, the glial cells, which surround the various parts of neurons and cooperate with them. Glial cells are discussed in Chapter 2.

Neurons are excitable cells. Depending on the information they receive, neurons generate electrical signals and propagate them along their processes. This capacity is due to the presence of particular proteins in their plasma membrane which allow the selective passage of ions: the ion channels.

Neurons are also secretory cells. Their secretory product is called a neurotransmitter. The release of a neurotransmitter occurs only in restricted regions, the synapses. The neurotransmitter is released in the extracellular space. The synaptic secretion is highly focalized and directed specifically on cell regions to which the neuron is connected. The synaptic secretion is then different (with only a few exceptions) from other secretory cells, such as from hormonal and exocrine cells which respectively release their secretory products into the general circulation (endocrine secretion) or the external environment (exocrine secretion). Synapses are discussed in Chapter 6.

Neurons are quiescent cells. When lesioned, most neurons cannot be replaced, since they are postmitotic cells. Thus, they renew their constituents during their entire life, involving the precise targeting of mRNAs and proteins to particular cytoplasmic domains or membrane areas.

1.1. Neurons have a cell body from which emerge two types of processes: the dendrites and the axon

Although neurons present varied morphologies, they all share features that identify them as neurons. The cell body or soma gives rise to processes which give the neuron the regionalization of its functions, its polarity and its capacity to connect to other neurons, to sensory cells or to effector cells.

1.1.1. The somatodendritic tree is the neuron’s receptive pole

The soma of the neuron contains the nucleus and its surrounding cytoplasm (or perikaryon). Its shape is variable: pyramidal soma for pyramidal cells in the cerebral cortex and hippocampus; ovoid soma for Purkinje cells in the cerebellar cortex; granular soma for small multipolar cells in the cerebral cortex, cerebellar cortex and hippocampus; fusiform soma for neurons in the pallidal complex; and stellar or multipolar soma for motoneurons in the spinal cord (Figure 1.1).

Figure 1.1   The neurons of the central nervous system present different dendritic arborizations.

(a) Photomicrographs of neurons in the central nervous system as observed under the light microscope. A – Purkinje cell of the cerebellar cortex; B – pyramidal cell of the hippocampus; C – soma of a motoneuron of the spinal cord. Golgi (A and B) and Nissl (C) staining. The Golgi technique is a silver staining which allows observation of dendrites, somas and axon emergence. The Nissl staining is a basophile staining which displays neuronal regions (soma and primary dendrites) containing Nissl bodies (parts of the rough endoplasmic reticulum). (b) Camera lucida drawings of neurons in the central nervous system of primates, revealed by the Golgi silver impregnation technique and reconstructed from serial sections: St, medium spiny neuron of the striatum; GP, neuron of the globus pallidus; Th, thalamocortical neuron; STN, neuron of the subthalamic nucleus; IO, neurons of the inferior olivary complex; Pu, Purkinje cell of the cerebellar cortex; SNc, dopaminergic neuron of the substantia nigra pars compacta. All these neurons are illustrated at the same magnification. Photomicrographs by Francoise Condé (aA), Olivier Robain (aB) and Paul Derer (aC). Drawings by Jérôme Yelnik, except OL and PU by Ramon Y Cajal (1911).

One function of the soma is to ensure the synthesis of many of the components required for the structure and function of a neuron. Indeed, the soma contains all the organelles responsible for the synthesis of macromolecules. Most neurons in the central nervous system cannot further divide or regenerate after birth, and the cell body must maintain the structural integrity of the neuron throughout the individual’s entire life. Moreover, the soma receives numerous synaptic contacts from other neurons and constitutes, with the dendrites, the main receptive area of neurons (see Figure 1.5 and Section 6.2). The neurons have one or several processes emerging from the cell body and arborizing more or less profusely. The two types of neuronal processes are the dendrites and the axon (Figures 1.1 and 1.3). This division is based on morphological, ultrastructural, biochemical and functional criteria.

The dendrites, when they emerge from the soma, are simple perikaryal extensions, the primary dendrites. On average, between one and nine primary dendrites emerge from the soma and then divide successively to give a dendritic tree with specific characteristics (number of branches, volume, etc.) for each neuronal population (Figures 1.1 and 1.2). The dendrites are morphologically distinguishable from axons by their irregular outline, by their diameter, which decreases along their branchings, by the acute angles between the branches, and by their ultrastructural characteristics (Figures 1.1, 1.3 and 1.7). The irregular outline of dendrites is related to the presence of numerous appendices of various shapes and dimensions at their surface. The most frequently observed are the dendritic spines which are lateral expansions with ovoid heads binding to the dendritic branches by a peduncle that is variable in length (Figure 1.3). Some neurons are termed ‘spiny’ because there are between 40 000 and 100 000 spines on the surface of their dendrites (e.g. pyramidal neurons of the cerebral cortex and hippocampus, the medium-sized neurons of the striatum, and the Purkinje cells of the cerebellar cortex). However, other neurons with only a few spines on their dendritic surface are termed ‘smooth’ (e.g. neurons of the pallidal complex) (Figure 1.1). The transition from the cell body to proximal dendrites is gradual, and the cytoplasmic architectures of proximal dendrites and the cell body are similar. In particular, the endoplasmic reticulum and ribosomes are almost as abundant in the proximal dendrites as in the cell body. Moreover, even distal dendrites contain ribosomes and endoplasmic reticula.

Figure 1.2   Tridimensional illustration of a dendritic arborization.

Computer drawing of a neuron of the subthalamic nucleus injected intracellularly with horseradish peroxidase (HRP) and reconstructed in three dimensions from serial sections. At 0°, the dendritic arborization of this neuron is represented in its principal plane; i.e. in the plane where it has its largest surface. In this plane, the dendritic field is almost circular (859 μm long and 804 μm wide). 30°, 60° and 90° rotations from the principal plane around the horizontal (horizontal column) and vertical (vertical column) axis show that the dendritic field has a flattened ovoidal form (230 μm thick). From Hammond C, Yelnik J (1983) Intracellular labelling of rat subthalamic nucleus with horseradish peroxidase: computer analysis of dendrites and characterization of axon arborization. Neuroscience 8, 781–790, with permission.

Figure 1.3   Dendritic spines and axon collaterals.

Dendritic spines of a pyramidal neuron of the hippocampus after 19 days of culture in vitro. Scale bar: 20 μm. (b) Axonal arborization of a Purkinje cell and enlargement of axonal bifurcation. Scale bars 50 and 10 μm. Part (b) photo by Hartmut Schmidt.

Dendrites and soma receive numerous synaptic contacts from other neurons and constitute the main receptive area of neurons (see Figure 1.5 and Section 6.2). In response to afferent information, they generate electrical signals such as postsynaptic potentials (EPSPs or IPSPs; see Figure 1.5 left 1, 2) or calcium action potentials, and integrate the afferent information. Chapters 8–10 look at the mechanisms underlying the excitatory (EPSP) and inhibitory (IPSP) postsynaptic potentials generated in the postsynaptic membrane in response to transmitter release. Chapters 13–16 discuss how these postsynaptic responses are integrated along the somatodendritic tree. Although dendrites are generally a receptive zone, there are certain exceptions: some dendrites are connected with other dendrites and act as a transmitter area by releasing neurotransmitters (see Figure 6.2d).

1.1.2. The axon and its collaterals are the neuron’s transmitter pole

The axon is morphologically distinct from dendrites in having a smooth appearance and a uniform diameter along its entire extent, and by its ultrastructural characteristics (see Figures 1.3, 1.6 and 1.7). Axons are narrow from their origin, and do not usually contain ribosomes or endoplasmic reticula. The transition from the cell body to axon is distinct; the region of the cell body from which an axon originates is called the axon hillock and it tapers off to the axonal initial segment, where action potentials begin. Although most parts of the cell body are rich in endoplasmic reticula, the axon hillock is not. At the axon initial segment, the plasma membrane has thick underlying structures, and there is a specialized bundle of microtubules. In some neurons, the axon emerges at the level of a primary dendrite.

The axon is not a single process; it is divided into one or several collaterals which form right-angles with the main axon. Some collaterals return toward the cell body area; these are recurrent axon collaterals. The axon and its collaterals may be surrounded by a sheath, the myelin sheath. Myelin is formed by glial cells (see Sections 2.2 and 2.4). The length of an axon varies. Certain neurons in the central nervous system have axons that project to one or several structures of the central nervous system that are more or less distant from their cell bodies (Figure 1.4), whereas other neurons have short axons (a few microns in length) that are confined to the structure where their cell bodies are located; these are interneurons or local circuit neurons (see Figure 1.13).

Figure 1.4   Neurons showing complex axonal arborizations.

Drawing of a cat reticulospinal neuron has been stained by intracellular injection of peroxidase and drawn in a parasagittal plane obtained from serial sections. The axon (ax, black) gives off numerous collaterals along its rostrocaudal trajectory, making contacts with different neuronal populations (delimited by broken lines). Scale: 7 mm = 1 mm. (b) A rat GABAergic ‘hub’ neuron of the CA3 region of the hippocampus filled with neurobiotin during whole cell recording shows numerous axonal collaterals (blue) that expand inside and outside the hippocampus. In comparison a control GABAergic interneuron shows a restricted axonal arborization (green). Cells bodies and dendrites are in black. Part (a) from Grantyn A (1987) Reticulo-spinal neurons participating in the control of synergic eye and head movement during orienting in the cat. Exp. Brain Res. 66, 355–377, with permission. Part (b) from Picardo MA, Guigue P et al. (2011) Pioneer GABA cells comprise a subpopulation of hub neurons in the developing hippocampus. Neuron 71, 695–709, with permission.

Thus projection (Golgi type I) neurons and local-circuit (Golgi type II) neurons can be differentiated. In Golgi type I neurons, the length of the axon is variable: certain projection neurons are directed to one structure only (e.g. corticothalamic neurons; see Figure 1.14) whereas other projection neurons have numerous axon collaterals which project to several cerebral structures (Figure 1.4).

The axon and axonal collaterals in certain neurons end in a terminal arborization, i.e. numerous thin branches whose extremities, the synaptic boutons, make synaptic contacts with target cells (see Figure 6.3). In other neurons, the axon and its collaterals have enlargements or varicosities which contact target cells along their way: these are ‘boutons en passant’ (see Figures 6.14 and 6.15b). It can be noted that both types of boutons are called axon terminals, although ‘boutons en passant’ are not the real endings of the axon.

The main characteristic of axons is their capacity to trigger sodium action potentials and to propagate them over considerable distances without any decrease in their amplitude (Figure 1.5 left 3). Action potentials are generated at the initial segment level in response to synaptic information transmitted by the somatodendritic tree. Then they propagate along the axon and its collaterals toward the axon terminals (synaptic boutons or boutons en passant). When action potentials reach the axon terminals these trigger calcium action potentials (Figure 1.5 left 4) which may cause the release of the neurotransmitter(s) contained in axon terminals in a specific compartment, the synaptic vesicles. This secretion is localized only at the synaptic contacts. Overall, the axon is considered as the transmitter pole of the neuron.

Figure 1.5   Comprehensive schematic drawing of neuron polarity.

The somatodendritic compartment of a neuron receives a large amount of information from other neurons that establish synapses with it. At each synapse level, the neuron generates postsynaptic potentials in response to the released neurotransmitter (1, EPSP; 2, IPSP). These postsynaptic potentials propagate and summate in the somatodendritic compartment, then they propagate to the initial segment of the axon where they generate (or not) action potential(s) (3a). The action potentials propagate along the axon (3b, 3c) and its collaterals up to the axon terminals where they evoke (or not) the entry of calcium (4) and neurotransmitter release. Note the different voltage and time calibrations.

Chapter 4 discusses the mechanisms underlying the abrupt, large and transient depolarizations called (sodium) action potentials, and how they are triggered and propogated. Chapters 4, 5 and 7 look at how sodium action potentials trigger calcium action potentials, the entry of calcium in synaptic terminals and the secretion of transmitter molecules.

Certain regions – such as the initial segment, nodes of Ranvier (zones between two myelinated segments; see Figure 1.5) and axon terminals – can also be receptive areas (a postsynaptic element) of synaptic contacts from other neurons (see Section 6.2).

1.2. Neurons are highly polarized cells with a differential distribution of organelles and proteins

The somatodendritic tree is the neuron’s receptive pole, whereas the axon and its collaterals are the neuron’s transmitter pole. Neurons are highly polarized cells. Cellular morphology and accurate organelles and protein distribution lay the basis to this polarization. The organelles and cytoplasmic elements present in neurons are the same organelles found in other cell types. However, some elements such as cytoskeletal elements are more abundant in neurons. The non-homogeneous distribution of organelles in their soma and processes is one of the most distinguishing characteristics of neurons.

1.2.1. The soma is the main site of macromolecule synthesis

The soma contains the same organelles and cytoplasmic elements that exist in other cells: cellular nucleus, Golgi apparatus, mitochondria, polysomes, cytoskeletal elements and lysosomes. The soma is the main site of synthesis of macromolecules since it is the one compartment containing all the required organelles.

Compared with other types of cells, the neuron differs at the nuclear level and more specifically at the chromatin and nucleolus levels. The chromatin is light and sparsely distributed: the nucleus is in interphase. Indeed, in humans, most neurons cannot divide after birth since they are postmitotic cells. The nucleolus is the site of ribosomal synthesis and ribosomes are essential for translating messenger RNA (mRNA) into proteins. The large size of the nucleolus indicates a high level of protein synthesis in neurons.

1.2.2. The dendrites contain free ribosomes and synthesize some of their proteins

In dendrites can be found smooth endoplasmic reticulum, elongated mitochondria, free ribosomes or polysomes, and cytoskeletal elements including microtubules which are oriented parallel to the long axis of the dendrites (Figure 1.6).

Figure 1.6   Photomicrograph of a tissue section of the central nervous system at the hippocampal level.

This shows the ultrastructure of a dendrite, numerous axons and their synaptic contacts (observation under the electron microscope). The apical dendrite of a pyramidal neuron contains mitochondria, microtubules, ribosomes and smooth endoplasmic reticulum. It is surrounded by fascicles of unmyelinated axons with mitochondria and microtubules but no ribosomes. The axon’s trajectory is perpendicular to the section plane. Three synaptic boutons (Ax Term) with synaptic vesicles make synaptic contacts (arrows) with the dendrite. Photomicrograph by Olivier Robain.

By using the hook procedure, microtubules have been shown to have two orientations in proximal dendrites: half of them are oriented with the plus-ends distal to the cell body, and the other half has the plus-ends proximal to the cell body. This is very different from the orientation in distal dendrites and axons (Figure 1.7), which is uniform. Moreover, one microtubule-associated protein (MAP), the high-molecular-weight MAP2 protein and more precisely the MAP2A and MAP2B, are more common to dendrites than to axons. For this reason MAP2A or MAP2B antibodies coupled to fluorescent molecules are useful for labeling dendrites, particularly for dendrite identification in cell cultures.

Figure 1.7   Microtubule polarity in neuronal processes.

(a) The polarity of microtubules is defined by the hook procedure. Neurons in culture are permeabilized in the presence of taxol to stabilize microtubules. Monomers of tubulin, purified from brain extracts, are added in the extracellular medium. Several minutes after, transversal cuts are performed at the level of dendrites or at the level of an axon. Slices are treated for electron microscopy. Hook-like structures are observed. They result from exogenous tubulin polymerization at the surface of endogenous microtubules. Hooks are always oriented toward the plus-end of microtubules. (b) When hooks, at the electron microscopic level, have mixed orientations (clockwise and anticlockwise), this means that endogenous microtubules are antiparallel (left). Uniformly oriented hooks (right) indicate that endogenous microtubules are parallel. (c) Orientation of microtubules in dendrites and the axon. Drawing (a) by Lotfi Ferhat. Drawing (b) adapted from Sharp DJ, Wenqian Yu, Ferhat L et al. (1997) Identification of a microtubule-associated motor protein essential for dendrite differentiation. J. Cell. Biol. 138, 833–843. Drawing (c) adapted from Baas PW, Deitch JS, Black MM, Banker GA (1988) Polarity orientation of microtubules in hippocampal neurons: uniformity in the axon and nonuniformity in the dendrite. Proc. Natl Acad. Sci. USA 85, 8335–8339, with permission.

mRNA trafficking and local protein synthesis in dendrites

The dendritic compartment contains many ribosomes whereas an axon has considerably fewer ribosomes. One particular feature of dendrites, compared with axons, is the presence of synapse-associated polyribosome complexes (SPRCs); these are clusters of polyribosomes and associated membranous cisterns that are selectively localized beneath synapses (more precisely, beneath postsynaptic sites), at the base of dendritic spines when spines are present.

What is the origin of this selective distribution of ribosomes in neurons? This question is particularly important since this compartmentalization leads to different properties of dendrites and axons: dendrites can locally synthesize some of their proteins, whereas axons would synthesize very few of them, if any.

Whereas most proteins destined for dendrites and dendritic spines are conveyed from the cell body, a subset of mRNAs is transported into dendrites to support local protein synthesis. Such a local dendritic protein synthesis requires that a particular subset of mRNAs synthesized in the nucleus is transported into the dendrites up to the polysomes where they are translated. How are dendritic mRNAs recognized and routed to dendrites and not axons? How are they transported to their final dendritic destination?

Chimeric constructs consisting of a reporter gene fused to putative cis-acting RNA localization elements were used to visually identify cis-acting sequences that direct mRNAs to their destination in dendrites near synapses. These experiments convincingly proved the existence of cis-acting elements commonly referred to as ‘zipcodes’ which allow mRNAs targeting to distinct subcellular compartments in order to regulate gene expression with temporal and spatial control.

In cultured hippocampal neurons, RNA labeled with tritiated uridine is shown to be transported at a rate of 250–500 μm per day. This transport is blocked by metabolic poisons and the RNA in transit appears to be bound to the cytoskeleton, since much of it remains following detergent extraction of the cells. Studies using video microscopy techniques and cell-permeant dyes which fluoresce on binding to nucleic acids have permitted observation of the movement of RNA-containing particles along microtubules in dendrites. These studies suggest that mRNAs are transported as part of a larger structure. The visualized RNA particles co-localize with poly(A) mRNA, the 60S ribosomal subunit, suggesting that the RNA-containing particles may represent translational units or complexes (Figure 1.8). This energy-dependent transport seems to be associated with the dendrite cytoskeleton as also shown by the delocalization of RNA-containing particles in response to colchicin (a drug which blocks microtubule polymerization).

Figure 1.8   Approximate sizes of representative dendritic mRNAs and translational elements at synaptic sites on dendrites.

The drawing illustrates the approximate size range of spine synapses that would be found in rat forebrain structures such as the hippocampus and cerebral cortex. The lines represent the approximate length of representative dendritic mRNAs if they were straightened out. Shading indicates the length and position of the coding region. Adapted from Schuman EM, Dynes JL, Steward O (2006) Synaptic regulation of dendritic mRNAs. J. Neurosci. 26, 7143–7146, with permission.

To visualize mRNA translocation in live neurons, studies used nucleic acid stains and green fluorescent protein fused to RNA-binding proteins. It showed that mRNAs are transported in the form of large particles, the ribonucleoprotein particles, containing mRNAs, RNA-binding proteins, ribosomes, and translational factors (RNA-containing granules) in a rapid (average speed, 0.1 μm/s), bidirectional, and microtubule dependent manner. Ribonucleoprotein particles that are transported to dendrites bind to motor proteins (kinesin) as a large detergent resistant, RNAse-sensitive particle (see Figure 1.11 and Section 1.3 for kinesin).

The mRNAs present in dendrites encode proteins with specialized synaptic functions. Among the mRNAs detected in dendrites by in situ hybridization (see Appendix 6.2) are mRNAs that encode certain cytoskeletal proteins (Arc, MAP2), kinases (the α-subunit of calcium/calmodulin-dependent protein kinase II), an integral membrane protein of the endoplasmic reticulum (the inositol trisphosphate receptor), calcium-binding proteins, certain subunits of neurotransmitter receptors (GluR1 and 2 of AMPA receptors, NR1 of NMDA receptors) as well as other proteins of unknown function. Moreover, within dendrites, different mRNAs are localized in different domains and different mRNAs are localized in the dendrites of different neuron types.

In summary, it has become clear that certain mRNAs are targeted to dendrites within ribonucleoprotein particles (>1000S). These particles are transported distally by a kinesin, which binds RNA-binding proteins by a recognition motif in its tail domain. mRNAs targeted to dendrites would travel to their final synaptic destination in a translationally repressed state due to the presence of various binding partners acting as translational regulators. Once an individual synapse is activated, these particles would partially disassemble and translational repression would be relieved locally.

The relatively large amount of RNA transported into dendrites raises the question of why neurons need this supply. Targeting of mRNAs to dendritic synthetic machinery located at the base of dendritic spines suggests that local translation could be regulated in an activity-dependent manner. For example, in response to synaptic information local protein synthesis is triggered to modulate synaptic transmission by changing, for example, the subunits or the number of receptors to the neurotransmitter in the postsynaptic membrane (this can occur during plasticity and may produce long-lasting changes in synaptic strength) (see Chapter 18). This in turn would allow an individual synapse of a given neuron to modify its function as well as its morphology, thereby providing a mechanism for synaptic plasticity and memory consolidation.

1.2.3. The axon, to a large extent, lacks the machinery for protein synthesis

The axoplasm contains thin elongated mitochondria, numerous cytoskeletal elements and transport vesicles. It is devoid of ribosomes associated to the reticulum but may contain ribonucleoprotein complexes especially during development. Nevertheless, axons cannot restore the vast majority of the macromolecules from which they are made; neither can they ensure alone the synthesis of the neurotransmitter(s) that they release since they are unable to synthesize proteins (such as enzymes). This problem is resolved by the existence of a continuous supply of macromolecules from the cell body to the axon through anterograde axonal transport (see Section 1.3).

Another major difference between dendrites and axons is the orientation of microtubules. By using the hook procedure (see Figure 1.7), it has been shown that the polarity of microtubules is uniform in the axon, meaning that all their plus-ends point away from the cell body, toward the axon terminals. The polarity of the microtubules is relevant for transport properties (see Section 1.3). Moreover, one MAP, the Tau protein, is more common to axons than to dendrites. Tau antibodies coupled to fluorescent molecules are useful for labeling axons, particularly for axon identification in cell cultures.

1.3. Axonal transport allows bidirectional communication between the cell body and the axon terminals

Axonal transport is the movement of subcellular structures (such as vesicles, mitochondria, etc.) and proteins (like those of the cytoskeleton) from the cell body to axonal sites (nodes of Ranvier, presynaptic release sites, etc.) and from axon terminals to the cell body.

1.3.1. Demonstration of axonal transport

Weiss and Hiscoe (1948) first demonstrated the existence of material transport in growing axons (during development) as in mature axons. Their work consisted of placing a ligature on the chicken sciatic nerve, and then examining the change in diameter of the axons over several weeks. They showed that these neurons became enlarged in their proximal part and presented degenerative signs in their distal part (Figure 1.9). The authors suggested that material from the cell body had accumulated above the ligature and ensured the survival of the distal part.

Figure 1.9   Experiment by Weiss and others demonstrating anterograde axonal transport.

Schematic of a chicken motoneuron (1). When a ligature is placed on the axon (2) an enlargement of the axon’s diameter above the ligature is noted after several weeks (3). When this ligature is removed, the enlargement progressively disappears (4). From Weiss P, Hiscoe HB (1948) Experiments on the mechanism of nerve growth. J. Exp. Zool. 107, 315–396, with permission.

Later, Lubinska et al. (1964) elaborated the concept of anterograde and retrograde transport. These authors placed two ligatures on a dog sciatic nerve, isolated part of the nerve and divided it into short segments in order to analyze their acetylcholinesterase content. This enzyme is responsible for acetylcholine degradation and was used here as a marker. They showed that it accumulates at the level of both ligatures. This result therefore suggested the existence of two types of transport: an anterograde transport (from cell body to terminals) and a retrograde transport (from terminals to cell body). Moreover, it appeared that both types of transport are distributed along the entire extent of the axon.

We presently know of three types of axonal transport: fast (anterograde and retrograde), slow (anterograde) and mitochondrial.

1.3.2. Fast anterograde axonal transport is responsible for the movement of membranous organelles from cell body towards axon terminals, and allows renewal of axonal proteins

Fast anterograde axonal transport consists in the movement of vesicles along the axonal microtubules at a rate of 200–400 mm per day (i.e. 2–5 μm/s). These transport vesicles, which are 40–60 nm in diameter, emerge from the Golgi apparatus in the cell body (Figure 1.10a). They transport, among other things, proteins required to renew plasma membrane and internal axonal membranes, neurotransmitter synthesis enzymes and neurotransmitter precursors when the neurotransmitter is a peptide. This transport is independent of the type of axon (central, peripheral, etc.).

Figure 1.10   Fast axonal transport.

(a) Schematic of fast anterograde axonal transport (anterograde movement of vesicles) and retrograde axonal transport (retrograde movement of plurivesicular bodies). These two transports use microtubules as substrate. The detail shows recycling of small synaptic vesicles. Vesicles synthetized in the cell body and transported to the axon terminals are loaded with cytoplasmic neurotransmitter and targeted to the presynaptic plasma membrane. In response to Ca²+ entry, they fuse with the plasma membrane and release their content into the synaptic cleft (exocytosis); then they are recycled via an endosomal compartment. (b) Schematic of mitochondrial transport. Note that the neuron representation is extremely schematic since axons do not give off one axon terminal. Drawing (a) adapted from Allen R (1987) Les trottoirs roulants de la cellule. Pour la Science, April, 52–66; and Südhof TC, Jahn R (1991) Proteins of synaptic vesicles involved in exocytosis and membrane recycling. Neuron 6, 665–677, with permission. Drawing (b) adapted from Lasek RJ, Katz M (1987) Mechanisms at the axon tip regulate metabolic processes critical to axonal elongation. Prog. Brain Res. 71, 49–60, with permission.

The pioneer living preparation

The squid’s giant axon is most commonly used for these observations since its axoplasm can easily be extruded and a translucent cylinder of axoplasm devoid of its membrane is thus obtained. This living extruded axon keeps its transport properties for several hours. The absence of plasma membrane allows precise control of the experimental conditions and entry into the axoplasm of several components that cannot usually pass through the membrane barrier in vivo (e.g. antibodies). The improvement of video techniques applied to light microscopy allowed the first observations of the movement of a multitude of small particles along the microtubules in a living extruded axon.

Identification of the moving organelles and their substrates

Analysis of the particles that accumulate on each side of the 1.0–1.5-mm long isolated frozen segments of the squid axon has permitted the identification of moving organelles in axons. Correlation between video and electron microscopy images of these axonal segments has shown that the particles moving anterogradely on video images are small vesicles. Indeed, when a purified fraction of small labeled vesicles (with fluorescent dyes) is placed in an extruded axon, these vesicles and also native vesicles are transported essentially in the anterograde direction.

Evidence demonstrating the implication of microtubules in fast anterograde transport came from experiments with antimitotic agents (colchicin, vinblastin) which prevent the elongation of microtubules and block this transport. Finally, video techniques have also demonstrated that the vesicles are associated to microtubules by arms of 25–30 nm length (Figure 1.11a).

Figure 1.11   The motors of fast anterograde transport.

(a) Kinesin motors (KIFs) carry cargoes (membranous organelles in the axon) anterogradely in axons along a unipolar array of microtubule towards the plus-ends. (b) Some members of kinesin superfamily proteins (KIFs) observed by low angle rotary shadowing (left column). Diagrams, constructed on the basis of electron microscopy or predicted from the analysis of their primary structures, are shown on the right (the larger orange ovals in each diagram indicate motor domains). Kinesin-1 consists of two KIF5s and two kinesin light chains (KLCs). The heavy chains form a homodimer that binds to the two KLCs to form a heterotetramer. KIF1A is a unique monomeric motor. KIF3 form a tetramer with two kinesin superfamily associated protein 3 (KAP3). (c) Simplified kinesin (K) mechanochemical cycle. ATP binding is thought to be the force producing step for kinesin motors. Part (a) and (b right) from Hirokawa N, Niwa S, Tanaka Y (2010) Molecular motors in neurons: transport mechanisms and roles in brain function, development, and disease. Neuron 68, 610–638, with permission. Part (b left) adapted from Hirokawa N (1998) Kinesin and dynein superfamily proteins and the mechanism of organelle transport. Science 279, 519–526. Part (c) adapted from Kull FJ, Endow SA (2013) Force generation by kinesin and myosin cytoskeletal motor proteins J. Cell Sci. 126, 9–19, with permission.

The role of ATP and kinesin

By analogy with actin–myosin movements in muscle cells, scientists tried to isolate in neurons an ATPase (the enzyme responsible for the hydrolysis of ATP) associated with microtubules and able to generate the movement of vesicles. To demonstrate molecular components responsible for interactions between vesicles and microtubules, the vesicle–microtubule complex system has been reconstituted in vitro: isolated vesicles from squid giant axons are added to a preparation of purified microtubules and placed on a glass coverslip. These vesicles occasionally move in the presence of ATP. If an extract of solubilized axoplasm is then added to this system, the number of transported vesicles is considerably increased.

In order to determine the factor present in the solubilized fraction responsible for vesicle movement, a non-hydrolyzable ATP analog has been used: the 5′-adenylyl imidophosphate (AMP-PNP). In the presence of AMP-PNP, the vesicles associate with the microtubules but then stop. In these conditions, vesicles are bound to the microtubules and also, consequently, to the transport factor. When an overdose of ATP is added to this vesicle–microtubule complex isolated by centrifugation, the AMP-PNP is removed and so vesicles are released and the transport factor is solubilized. Kinesin has been thus isolated and purified. It is a soluble microtubule-associated ATPase (MAP) belonging to the family of mechanochemical ATPases that couple ATP hydrolysis to unidirectional movement of vesicles along the microtubule. As we have already seen, in axons, all microtubules are oriented, their plus-end being distally located from the cell body. It has been shown that kinesin, now named conventional kinesin (KIF5), moves vesicles in one direction only: from the minus-end toward the plus-end. All these results show that kinesin is responsible for anterograde transport.

In mammals, kinesin is a homodimer composed of two identical heavy chains associated with two light chains. These form an 80 nm rod-like molecule. Kinesin is composed of a motor domain (head, formed by the heavy chains), a stalk domain and a tail region (formed by the light chains). The motor domain binds to microtubules and moves on them by hydrolyzing ATP, whereas the tail regions recognize and bind to the cargo(s) (Figure 1.11a,b,c). The head binds to and dissociates from a microtubule through a cycle of ATP hydrolysis (Figure 1.11c). In proposed mechanism models, the arms observed between vesicles and microtubules in vitro would be kinesin.

The effects of mutations of the kinesin heavy-chain gene (khc) on the physiology and ultrastructure of Drosophila larval neurons have been studied. Motoneuron activity and corresponding synaptic (junctional) excitatory potentials of the muscle cells they innervate were recorded in control and mutant larvae in response to segmental nerve stimulation. The mutations dramatically reduced the evoked motoneuron activity and synaptic responses. The synaptic responses were reduced even when the terminals were directly stimulated. However, there was no apparent effect on the number of axons in the nerve bundle or the number of synaptic vesicles in the nerve terminal cytoplasm. These observations show that kinesin mutations impair the function of action potential propagation and neurotransmitter release at nerve terminals. Thus kinesin appears to be required for axonal transport of material other than synaptic vesicles: for example, vesicles containing ion channels such as Na+ channels delivered to Ranvier nodes and Ca²+ channels delivered to presynaptic membranes. These vesicles, called ‘cargoes’, are linked to kinesin. The observation that mutation of kinesin heavy chain had no effect on the number of synaptic vesicles within nerve terminals would obviously not be expected if conventional kinesin were the universal anterograde axonal transport motor.

Plus-end vesicle motors

Since the original discovery of kinesin, a large family of proteins (kinesin superfamily proteins or KIFs) with homology to kinesin’s motor domain has been discovered. The kinesin superfamily is a large gene family of microtubule-dependent motors with more than 40 members identified in mice and humans. The murine and human KIF genes have been classified into three types on the basis of the positions of their motor domains: N-terminal motor domain KIFs (N-KIFs), middle motor domain KIFs (M-KIFs) and C-terminal motor domain KIFs (C-KIFs). All KIFs have a globular motor domain that shows high degrees of homology and contains a microtubule-binding sequence and an ATP-binding sequence but, outside the motor domain, each KIF has a unique sequence (Figure 1.11a). Of these, only N-terminal KIFs generally move towards microtubule plus-ends (from the soma to the axon terminal). Kinesins have either a monomeric (ex KIF1A), a homodimeric (ex KIF5) or a heterodimeric (ex KIF3) structure (Figure 1.11b). The ‘classical kinesin’ corresponds to KIF5. Many KIFs are expressed primarily in the nervous system, but KIFs are also expressed in other tissues and participate in various types of intracellular transport. The diversity of cargo-binding domains explains how KIFs can transport numerous different cargoes. The hypothesis is that each KIF member is targeted to a specific cargo population, allowing the trafficking of the different neuronal compartments to be regulated independently. While there is some functional redundancy among members of the kinesin superfamily, there is also a remarkable degree of cargo specialization. For example, KIF1A/KIF1Bβ transport synaptic vesicle precursors, KIF5 transport presynaptic membrane or active zone vesicles and KIF1Bα/KIF5 transport mitochondria, all down the axon (Figure 1.11a).

1.3.3. Retrograde axonal transport returns old membrane constituents, trophic factors, exogenous material to the cell body

Retrograde axonal transport allows debris elimination and could represent a feedback mechanism for controlling the metabolic activity of the soma. The vesicles or cargoes transported retrogradely are larger (100–300 nm) than those transported anterogradely. Structurally, they are prelysosomal structures, multivesicular or multilamellar bodies (Figure 1.10). In the squid extruded axoplasm, vesicles move on each filament in both directions and frequently cross each other without apparent collisions or interactions.

Do filaments used for the fast transport of vesicles form a complex made up of several distinct filaments where certain filaments would be implicated in fast anterograde and others in retrograde transport? By using a monoclonal antibody raised against α-tubulin (a specific component of microtubules), it has been demonstrated that all the filaments implicated in anterograde or retrograde axonal transport contain α-tubulin. Moreover, by using a toxin-binding actin (and so consequently binding microfilaments), it was shown that filaments used for fast anterograde transport or retrograde transport were devoid of actin in their structure. Thus it appeared that filaments used for the movement of vesicles in both directions are microtubules.

The minus-end motor(s)

Morphometric analysis of the arms between retrograde vesicles (pluricellular bodies) and microtubules demonstrated that these are similar to arms between anterograde vesicles and microtubules. Studies looking to find a factor different from, but homologous to, kinesin and responsible for retrograde transport were undertaken. This factor present in axoplasm homogenate might be lost during kinesin purification procedures since no retrograde vesicles movement was observed in vitro with kinesin. Cytoplasmic dynein has been thus isolated. It is a microtubule-associated protein with an ATPase activity. Dynein belongs to the AAA+ (ATPase associated with diverse cellular activities) family of proteins.

Dynein is unique compared with kinesin and myosin because dynein molecules form large molecular complexes. Mammalian cytoplasmic dynein consists of two heavy chains, so termed because of their large molecular mass (each around 500 kDa), and several smaller subunits. Each heavy chain is divided into four domains: tail, linker, head, and stalk (Figure 1.12a,b). The tail is the cargo binding domain, the head is the motor domain (site of ATP hydrolysis), the linker is the mechanical amplifier, and the stalk is the microtubule binding domain. The divergent amino-terminal tail domain serves as a platform for the binding of several types of associated subunit which, in turn, mediate interactions with cargo either via direct binding or through the recruitment of adaptor proteins. Cytoplasmic dynein assembles around two identical heavy chains and is thus known as a two-headed motor.

Figure 1.12   Cytoplasmic dynein drives retrograde axonal transport in motor neurons.

(a) The primary structure of cytoplasmic dynein. (b) Schematic domain structure of dynein. (c) The cytoplasmic dynein complex contains a pair of identical heavy chains. Within each heavy chain, the six AAA+ modules fold into a ring. The stalk protrudes as an extension from the small subdomain of AAA4. The tail is connected to AAA1 by the linker domain, which arches over the AAA+ ring. The cytoplasmic dynein heavy chains assemble with up to five types of associated subunit, which are also dimers. Part (a) and (b) © Kikkawa 2013. Originally published in J. Cell Biology. 202:15–23. doi:10.1083/jcb.201304099. Part (c) adapted from Roberts AJ, Kon T, Knight PJ, Sutoh K, Burgess SA (2013) Functions and mechanics of dynein motor proteins. Nat. Rev. Mol. Cell Biol. 14, 713–726, with permission.

The motor domain has been uncovered through sequence analysis, two-dimensional (2D) electron microscopy and, most recently, X-ray crystallography and three-dimensional (3D) cryo-electron microscopy (cryo-EM). Like conventional AAA+ ATPases, dynein has a ring of six AAA+ modules at its core. The AAA+ ring can be thought of as the engine of dynein as it converts the chemical energy from ATP hydrolysis into motion.

Functions of retrograde transport

The removal of misfolded or aggregated protein is a key problem in neurons. Cytoplasmic dynein’s role as a retrograde motor makes it an ideal candidate for ‘taking out the trash’ in the cell, returning misfolded or degraded proteins from the cell periphery to the cell center for recycling and/or degradation. Evidence for such a role has come from the analysis of aggresome formation, in which the formation of perinuclear aggregates of misfolded protein was found to be dynein-dependent. The organelles trafficked by cytoplasmic dynein include endosomes, lysosomes, phagosomes, melanosomes, peroxisomes, mitochondria and vesicles from the endoplasmic reticulum (ER) destined for the Golgi. Retrograde axonal transport allows the return of membrane molecules to cell bodies, where they are degraded by acidic hydrolases found in lysosomes.

Retrograde axonal transport is not only a means of transporting cellular debris for their elimination, but is also a way of communicating information from the axon terminals to the soma. The retrogradely transported molecules would inform the cell body about activities taking place at the axon terminal level, or they may even have a neurotrophic action on the neuron. One key role for cytoplasmic dynein in neurons is in retrograde signaling, specifically the transport of neurotrophic factors from synapse to cell body. Neurotrophins are a family of small molecules, such as the nerve growth factor (NGF), the brain-derived neurotrophic factor (BDNF), and the neurotrophic factor NT3, which are secreted by target tissues, and then bind to receptor tyrosine kinases (Trk receptors) on the surface of the neuron. The neurotrophin/Trk receptor complex is then internalized (taken up by endocytosis) transported to the cell body where it initiates signaling cascades that regulate cell growth and survival.

Furthermore, viruses such as HIV, herpes virus and adenovirus have evolved mechanisms to hijack cytoplasmic dynein to reach the nucleus. Tetanus toxin or cholera toxin macromolecules are taken up by axon terminals and have a toxic effect on the cell body. These toxins, as well as horseradish peroxidase (HRP), an enzyme taken up by the axon terminals, are used in research studies for the retrograde labeling of neuronal pathways.

In conclusion, cargoes are transported in either the antero- or retrograde direction, depending on whether plus- or minus-end motors are active on their surface. Cargoes destined for the nerve terminal, such as synaptic vesicles or their precursors, are transported by plus-end motors; while cargoes targeted for the cell body, such as vesicles containing neurotrophin-receptor complexes, are transported by minus-end motors. In axons, oriented microtubules establish a ‘road map’ inside the neuron to motors that are linked to particular intracellular cargoes.

1.3.4. Slow anterograde axonal transport moves cytoskeletal proteins and cytosoluble proteins

The cytoskeleton (microtubules, neurofilaments and microfilaments) and cytosolic proteins (intermediate metabolic enzymes including glycolysis enzymes) are transported anterogradely along axons at a slow rate of about 0.002–0.1 μm/s (0.17–8.6 mm/day). In the elongating axon (i.e. during development or regeneration), the function of the slow transport is to supply axoplasm required for axonal growth. In mature neurons, its function is to renew continuously the total proteins present in the axon and axon terminals and to act as a substrate for the anterograde and retrograde axonal transport. To appreciate fully the structural achievement of this transport, one must put the size of cell bodies and axons into relation. The neuronal cell bodies (10–50 μm diameter) are connected by axons that can be over 1 m length (the axonal diameter is 1 to 25 μm). This is a factor of 100 000 difference. As there is relatively little protein synthesis in the axon, the proteins that comprise the microtubules, neurofilaments and microfilaments must be actively transported from the cell body into and down the length of the axon.

To understand the mechanisms involved in slow axonal transport, several questions can be raised: (i) in which state are cytoskeletal proteins transported in the axons: as soluble proteins or as polymers? (ii) in which axonal region(s) is the cytoskeleton (i.e. the complex network of filaments) assembled? The following are, in chronological order, the diverse hypotheses that